Table of Contents

  • Your data.Your AI or your app.
  • Across your portfolio.
  • One real estate foundation.
  • Ask in AI. Build an app.
  • Your AI.On your terms.
  • Independence.
  • Automation.
  • Ownership.
  • Your data connects.Your company context carries through.
  • Own your AI.Don’t rent it.
  • Use your AI or build an app.
  • Keep what your team learns.
  • Build toward your own models.
  • “Which units took longest to turn?”
  • Give property teams a turn board.
  • Monday startswith the work ready.
  • Know where to act.
  • Ask the business.Follow the evidence.
  • Close the books.The reports follow.
  • One company.Views that fit the job.
  • The lease. The model.The answer.
  • Already in motion.
  • Your branded apps
  • Your portfolio, every morning
  • Recurring work, delivered
  • One foundation, every interface
  • AI where the team works
  • See adoption and usage
  • Your next answer.Your experience, built in.
  • Which renewals need attention?
  • The decision.And how you got there.
  • Connector-Routed Knowledge Graph
  • Chunk-and-vector baseline
  • Leni RKG
  • Universal Data Model
  • Your warehouse, ready for AI.
  • Use it in your AI or app.
  • Adapters for different systems
  • Validation and anomaly checks
  • Full lineage and versioning
  • History through system changes
  • The model proposes.The system checks.
  • Plug into your business.Build what makes it yours.
  • Your data.Your control.
  • SOC 2 Type II
  • Your deployment
  • Source permissions
  • Training rights, agreed
  • Full lineage and audit
  • Freedom to change
  • Your project.Managed by Leni.
  • A working app for your team.
  • The team thatbrings all three.
  • We've run real estate.
  • We build the AI.
  • We deliver inside enterprises.
  • A first result.Then a company system.
  • Trusted numbers.
  • Reporting on repeat.
  • Across the firm.
  • Build around the work.
  • Your OS.In leaders’ hands.
  • Connect the first assets.
  • Review in private.
  • Use it to decide.
  • The work you do today.The AI you own tomorrow.
Own your AI

Your company.
Your AI operating system.

Use your portfolio data in Claude or ChatGPT, or build a custom app for your business. Bring what you’ve already built. Leni can run the infrastructure behind it.

Connect your data ↗See customer stories →

Grow income. Reduce the cost of work. Keep what makes you different.

YOUR CHOICE OF MODELS
AnthropicOpenAIYour models⇄
Models change. Your business stays yours.
Your AI operating systemwith Leni
OverviewPortfolioReporting
Portfolio reportsYour numbers. Your format.
Deal underwritingYour investment rules.
↗
Connected dataOne set of definitions
Company knowledgeYour rules and decisions
Approved work becomes an asset you keep
Your knowledge. Your future AI.Agent traces → private training data
Run today. Build toward ownership.
Two ways to start

Your data.
Your AI or your app.

Start with Leni’s prebuilt real estate data warehouse or connect the warehouse you already have. Connect your assistant through MCP, or build an app on the same data through the API. Both use your definitions and permissions.

01 · Connect your sources

Across your portfolio.

YardiRealPageEntrataResManRent ManagerExcel + manager reports

Start with the sources your workflow needs.

02 · Leni data warehouse

One real estate foundation.

Connected portfolio dataConsistent definitionsHistorical records + source links

Built on the Universal Data Model.

03 · Choose your interface

Ask in AI. Build an app.

Claude or ChatGPTYour custom app

Ask which properties are over budget, or put the same figures in your asset managers’ dashboard.

Less gathering. More deciding.Reduce repeated downloads and spreadsheet assembly.
Add assets, keep the workflow.Bring new properties and managers into the same data model.
Keep the history useful.Plan historical data continuity when source systems change.

Start with one portfolio or workflow. We confirm the sources, history, refresh schedule and access your team needs.

See the data foundation ↓
The ownership advantage

Your AI.
On your terms.

Independence. Automation. Ownership. One foundation for the way your business works.

Anthropic⇄OpenAI
Your workflows stay

Independence.

Work in your preferred AI. Choose models for each task while keeping your data foundation, company knowledge and controls.

Keep your options open ↓
ConnectRun + checkDeliver
A complete workflow

Automation.

Move from prompts to repeatable work. Connect sources, run checks and deliver reports and apps, with people approving what matters.

See the work move ↓
ActionsCorrections
Your know-how, retained

Ownership.

Keep approved agent traces, decisions and corrections. Build a record of how your business works, ready for future model training.

See the long game ↓
The foundation beneath all three

Your data connects.
Your company context carries through.

The model gets the right evidence and business rules for the task. Approved work becomes knowledge you can reuse.

Your company knowledgeConnector-Routed Knowledge GraphDefinitions, source links and permissions.Your prebuilt data warehouseUniversal Data ModelConnected real estate data, available through MCP and API.
The long game

Own your AI.
Don’t rent it.

Start with one project. The apps, rules and decisions you build around it can serve the next one.

01 · START

Use your AI or build an app.

Ask a question in your usual AI, build a new app or bring a prototype. Start with the same approved property data.

Your data. Your interface.
02 · CAPTURE

Keep what your team learns.

Save the sources, checks and corrections behind each approved decision. Reuse them the next time the same question comes up.

Your private training set
03 · OWN

Build toward your own models.

Test a specialist model on work your team has approved. Move to private cloud or your own compute when the results justify it.

Your models. Your choice of infrastructure.
Each step should earn its place: better answers, less rework or a lower cost to run.
In your AI assistant

“Which units took longest to turn?”

Compare work-order dates with rent-ready dates. Check the source records before deciding where to intervene.

In your own app

Give property teams a turn board.

Show the same units, the next task and who owns it. Track days to rent-ready to see whether the process improves.

100,000+decision traces used in Leni’s specialist model training
~⅓of frontier-model cost in evaluated specialist workloads

Company-reported results. Your rollout and savings depend on the work.

What you keep, and what a trace contains

You retain your source data and agreed exports of company context, approved outputs and decision records. The build defines ownership and access for custom apps, workflows, training datasets and any models trained for you.

An agent trace records the sources used, tool calls, results, checks and human decisions. It can include explanations deliberately recorded by the workflow. It does not assume access to a model provider’s hidden internal reasoning.

Personal context stays separate from approved company knowledge. Only authorized, reviewed work enters a training set. The agreement defines reuse restrictions, retention and export formats.

A business with AI at work

Monday starts
with the work ready.

The numbers, answers and reports your team needs, ready for review.

Protect income

Know where to act.

See what changed. Prioritize the income at stake.

  • Renewals and concessions affecting NOI
  • Occupancy or collections falling behind plan
  • Market changes that put the next quarter at risk

A morning brief in Teams. An inbox with owners and drill-down.

Existing Leni financial performance visual with trends and an explanation
Decide with the same numbers

Ask the business.
Follow the evidence.

“How did the portfolio do against budget, and why?” Ask in plain language and trace the answer to the close.

  • Shared financial definitions across apps and models
  • Ledger, budget and property sources within reach
  • Access based on the person asking

Teams, Slack, Claude or ChatGPT. One governed foundation.

Leni answering a portfolio cash-flow question with an Excel output
Existing Leni interface · sample portfolio data
Give analysts their time back

Close the books.
The reports follow.

Accounting and manager data become your recurring pack.

  • Board packs, asset scorecards and investor updates
  • PDF, Word, Excel and PowerPoint outputs
  • Scheduled delivery to Teams, Outlook or SharePoint

Your formats. Your review rules. Less manual assembly.

↧Month-end close

Accounting and manager data arrive.

✓Checked and reconciled

Tied to the ledger. Exceptions go to review.

Board pack · PowerPoint
Portfolio review
Asset scorecards · Excel
Asset performance
Investor report · PDF
Fund update
Your brand, for every audience

One company.
Views that fit the job.

Your brand. The right view for each audience.

  • Executive NOI, occupancy, cash flow and debt views
  • Asset drill-down, statements and watch lists
  • Investor reports, capital accounts and documents

Draft privately. Review with your team. Publish to approved users.

Existing Leni financial dashboard used as the starting point for a company-branded executive view
Existing Leni dashboard · branding, audience and app scope configured together
Find the opportunity in the files

The lease. The model.
The answer.

Ask which leases roll below market, compare clauses or test a model assumption alongside the financial data.

  • Scanned PDFs, leases, contracts, email and Excel models
  • Clause and cell references for review
  • Source permissions carried into supported connections

SharePoint knowledge becomes part of the work.

Existing Leni interface comparing lease terms and flagging missing source text

Example working day built around existing Leni capabilities. Connections, schedules and approval rules are configured for your organization.

Already in motion.

Product walkthroughs using Leni’s existing interface. Availability and configuration depend on your organization.

Your branded apps

Your portfolio, every morning

Recurring work, delivered

One foundation, every interface

AI where the team works

See adoption and usage

Your advantage carries forward

Your next answer.
Your experience, built in.

Keep the investment logic and operating decisions that make your business different.

THE WORK

Which renewals need attention?

Lease dataMarket rentsYour NOI rules
THE REVIEW
Evidence + recommendation
Your team corrects and approves
WHAT YOU KEEP

The decision.
And how you got there.

Sources · actions · recorded reasoning · review

↶Approved lessons inform the next workflow↵

Only reviewed, authorized work becomes reusable company knowledge.

Patent pending

Connector-Routed Knowledge Graph

Your company’s memory. Connected to the source.

Your definitions and decisions guide each task. Evidence comes from the source, with the user’s permissions.

How retrieval compares

Chunk-and-vector baseline

  • Evidence copied into a separate index
  • Copies and permissions need to stay in sync
  • Retrieval can include loosely related text

Leni RKG

  • Evidence fetched from the connected source
  • Under the user's current source permissions
  • Only the context needed for the task
72% → 92%top-k retrieval accuracy
−61%context-token consumption
−42%median retrieval latency
<1 in 1,000stale or unauthorized retrievals

Company-reported internal evaluation against a chunk-and-vector baseline, with the base model held constant. Results describe the evaluated workload.

Patent pending

Universal Data Model

Your prebuilt real estate data warehouse.

Leni brings supported source data into one real estate model. Use consistent numbers across managers and systems, then access them from your AI, reports or apps.

7major ERPs supported
~85%of the relevant market
100M+rows processed daily

Company-reported platform figures. Specific systems, objects and permissions are confirmed for your implementation.

Yardi4010 · Residential Rent
Manager systemBase Rent Income
Excel fund modelGross Rent
Shared real estate modelRental revenue ·
Residential
Your definitionNOI, by your rules

Your warehouse, ready for AI.

Portfolio data, history and your definitions, organized by Leni.

MCP →

Use it in your AI or app.

Connect a compatible assistant through MCP or give your custom app access through the HTTP API.

How MCP connects your AI to Leni

MCP stands for Model Context Protocol. It lets a compatible AI assistant call Leni’s tools for the data and analysis a task needs. Each user signs in to Leni and approves access within their organization and permissions.

The warehouse holds the shared data foundation. MCP is the connection to it through Leni’s tools. Source coverage, history and refresh cadence are scoped with your team.

Read the MCP guide ↗

The real estate structure underneath
Fund›Entity›Property›Unit›Lease›Tenant
GL account›Account tree›Budget›Forecast

Adapters for different systems

Reconcile supported ERP structures into one model.

Validation and anomaly checks

Flag inconsistent data before analysis.

Full lineage and versioning

Trace transformed numbers back to their sources.

History through system changes

Scope historical imports and maintain mappings as sources evolve.

BUILT FOR TRUSTED OUTPUT

The model proposes.
The system checks.

  1. Run the task
  2. Check the sources
  3. Reconcile + verify
  4. Review + approve

Deterministic checks and a separate observation step sit around the model.

Specialist models and evaluation results

100,000+ decision traces across 30,000+ interactions inform Leni’s specialized models. Company-reported evaluations show 99.6%+ task accuracy and about 65% lower cost to reach the evaluated result. Results depend on the workload and baseline.

Your approved work can become a private training set. Model training and deployment follow the scope, rights and evaluations agreed with your team.

The system around the models

Plug into your business.
Build what makes it yours.

YOUR EXISTING SYSTEMSYardi · RealPage · MRISharePoint · ExcelSnowflake · market data
→
READY WITH LENI
Prebuilt real estate warehouseUDM
Company knowledgeRKG
Permissions + checks
Model-agnostic runtime
→
YOUR CHOICE OF EXPERIENCEClaude + compatible AIYour apps + agentsReports + decisionsAPI + MCP
Teams · Slack · Excel · company appsOne foundation. Your choice of interface.
One foundation, through the tools you use

Yardi, third-party managers, Microsoft 365, SharePoint, Snowflake, models and market data connect into the same governed foundation. The existing connector estate includes 1,500+ managed connections; confirm the specific sources in discovery.

A semantic layer gives metrics their business meaning. The runtime coordinates permitted tools, context and checks. Developers connect through the HTTP API; compatible assistants connect through MCP with Leni OAuth and the user's approved scope.

Teams, Slack, Outlook, Excel, Claude, ChatGPT, Copilot and company apps use the appropriate supported integration. The UDM and RKG are two patent-pending mechanisms within this wider system.

For developers & IT ↗Explore the architecture ↗
Control, built into the plan

Your data.
Your control.

Request security docs ↗

SOC 2 Type II

Independent examination. Report scope and period available for review.

Your deployment

Agree the environment, residency and sandbox boundaries with IT.

Source permissions

Access follows the user’s permissions in supported connections.

Training rights, agreed

Define private training, derived-data use and retention in your contract.

Full lineage and audit

Trace sources, changes, decisions and agent actions.

Freedom to change

Keep company context useful as models and systems evolve.

Security evidence and deployment choices

Leni has completed a SOC 2 Type II examination. The public report reference period is February 7 to May 7, 2024. Request the latest issued report for its scope and period. Data is encrypted in transit and at rest.

Agree private-cloud or customer-environment deployment, residency, source access, retention, model providers and operational responsibilities. Source permissions and isolation must be tested in the selected connections and deployment.

Leni’s current published terms permit some de-identified and derived data use for training and other purposes. A separate customer agreement defines any additional restrictions for your ownership plan.

Security and data use ↗Review the business and IT case →
Build with Leni

Your project.
Managed by Leni.

Bring an idea, a spreadsheet workflow or a working prototype. We review it with your team, keep what works and agree what it needs to go live.

WHAT LENI CAN MANAGE
Data connectionsAccess + deploymentMonitoring + support
+
WHAT YOUR TEAM DIRECTS
The business rulesYour app and brandWhat to build next

A working app for your team.

Your users see your product. Leni runs the services you choose.
From pilot to daily use.

Pilot with one team. Leni can maintain the connections, resolve failures and ship agreed changes as more properties come on.

Talk through your project →
How we measure the return

Agree the first workflow, data readiness and delivery milestones together. Track time to a trusted output, the cost per approved result and the income opportunities acted on.

Use approved agent traces to test and improve specialist models. Expand only after the workflow meets your team's quality, security and ownership requirements.

Real estate experience. Working AI.

The team that
brings all three.

We've run real estate.

Marcio Sahade brings 14 years at Hines and Tishman Speyer across asset management, acquisitions and capital raising.

We build the AI.

The data model, company knowledge graph and agent runtime. Two patent-pending foundations, with specialized models trained on real decision traces.

We deliver inside enterprises.

From scoping to adoption, alongside your sponsor, data owners and IT team. An existing platform serving 30+ organizations.

Meet the team ↗
Real estate
operatorsAI engineersEnterprise deliveryLeni
One team, from the investment decision to the working system.
$80B+in assets supported
30+organizations
1,000+users

Company-reported platform scale from the supplied reference materials.

Daily InsightsFinancial dashboardsTeams + SlackScheduled reportsBranded appsModel connections

Your build starts with an existing platform. Connection, approval and app scope are agreed for your organization.

Friedman’s reporting story →More customer stories →
Built with you

A first result.
Then a company system.

A shared roadmap connects the executive pilot to a wider rollout.

FIRST 90 DAYS

Trusted numbers.

Connect the first assets. Put an executive app and recurring report into use.

BY MONTH 6

Reporting on repeat.

Move approved reporting workflows, connect documents and models, and extend daily insights.

BY MONTH 12

Across the firm.

Expand team agents and investor views. Keep approved decisions and traces in your company context.

Timing follows scope and readiness. Business owners guide decisions; IT reviews access; Leni builds alongside your team.

What the leadership team measures

One set of reconciled numbers across reports, apps and answers. Less report assembly. Earlier action on income and operating issues. Fewer overlapping tools where consolidation makes sense. Company context and training data that remain usable as models or property systems change.

Model and source-system changes still need compatibility checks and mapping updates. A stable company data and context layer reduces the work of rebuilding applications around each change.

Build around the work.

Choose the workflow with a clear owner and a measurable business return.

Reporting & asset management↗Financial modeling & underwriting↗Investment memos & presentations↗Market research↗Document extraction↗Development & strategy↗
AppFolio ↗RealPage ↗Rent Manager ↗
The executive pilot

Your OS.
In leaders’ hands.

Use your own data in real meetings before deciding on the wider investment.

Your assistantQuestions, files and a Monday brief in Teams.
Your executive appNOI, occupancy and cash flow, with property drill-down.
Your investor viewApproved KPIs, reports and documents under your brand.
WEEKS 1–2

Connect the first assets.

Agree the scope and select an initial 5–10 buildings. Map the data and reconciliation rules.

WEEKS 3–4

Review in private.

Draft the apps. Your business and IT owners check the numbers, access and outputs.

WEEKS 5–6

Use it to decide.

Leaders use the apps in real meetings. Review adoption, reconciled numbers and reporting time saved.

The work you do today.
The AI you own tomorrow.

Bring a portfolio question or an app you’ve started. Agree what to pilot, who will run it and how you’ll measure the result.

Plan your executive pilot ↗